Generative AI as a Cybercrime Assistant

Anthropic reported that cybercriminals used the Claude AI model to automate reconnaissance, credential harvesting, network intrusion, data exfiltration, and psychologically targeted extortion across at least 17 organizations (including healthcare and emergency services). The actor used AI to choose what data to steal and to determine ransom amounts and messaging, and Anthropic disrupted the activity; the post also notes other AI-enabled misuse cases, such as remote-worker fraud linked to North Korean actors.